Natural Language Processing Engine Research Articles

Clinical Decision Support and Natural Language Processing in Medicine: Systematic Literature Review.

Ensuring access to accurate and verified information is essential for effective patient treatment and diagnosis. Although health workers rely on the internet for clinical data, there is a need for a more streamlined approach. This systematic review aims to assess the current state of artificial intelligence (AI) and natural language processing (NLP) techniques in health care to identify their potential use in electronic health records and automated information searches. A search was conducted in the PubMed, Embase, ScienceDirect, Scopus, and Web of Science online databases for articles published between January 2000 and April 2023. The only inclusion criteria were (1) original research articles and studies on the application of AI-based medical clinical decision support using NLP techniques and (2) publications in English. A Critical Appraisal Skills Programme tool was used to assess the quality of the studies. The search yielded 707 articles, from which 26 studies were included (24 original articles and 2 systematic reviews). Of the evaluated articles, 21 (81%) explained the use of NLP as a source of data collection, 18 (69%) used electronic health records as a data source, and a further 8 (31%) were based on clinical data. Only 5 (19%) of the articles showed the use of combined strategies for NLP to obtain clinical data. In total, 16 (62%) articles presented stand-alone data review algorithms. Other studies (n=9, 35%) showed that the clinical decision support system alternative was also a way of displaying the information obtained for immediate clinical use. The use of NLP engines can effectively improve clinical decision systems' accuracy, while biphasic tools combining AI algorithms and human criteria may optimize clinical diagnosis and treatment flows. PROSPERO CRD42022373386; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=373386.

Medication Recommendation System Using AI

Abstract: In an era marked by the convergence of personal and technical facets of life, the need for a versatile and intelligent companion becomes increasingly apparent. This project introduces the development of a novel Chatbot, aptly named the "Personal and Technical Companion," designed to seamlessly integrate into users' lives by offering personalized support and technical assistance. The Chatbot employs state-of-the-art Natural Language Processing (NLP) techniques to engage users in natural language conversations, providing a unique blend of personal and technical guidance. The system architecture encompasses a user-friendly interface, a robust NLP engine for intent recognition, a dynamic backend server handling logic and personalization, and integration with external services to enrich the user experience. Simultaneously, it serves as a technical companion by offering code assistance, troubleshooting guidance, and recommending relevant learning resources. Implementation leverages cutting-edge technologies, including Python for backend logic, Flask for the server, TensorFlow for NLP processing, and HTML/CSS for the user interface.

Connecting the dots with Hum: Unlocking data potential for publishers

This paper is based on a presentation by Dustin Smith at the APE 2023 Conference. Publishers struggle to gain a single view of audiences. As a consequence, they extract insights from inconsistent data sources. A customer data platform (CDP) and natural language processing engine help publishers solve these challenges. By integrating data, analyzing content, and using machine learning to understand individuals, publishers can personalize experiences, support key business cases, and tap into data’s full potential.

A KNOWLEDGE-DRIVEN BIO-BEHAVIORAL NETWORK MODEL OF THE INFLAMMASOME AND CALORIC RESTRICTION

Abstract Healthy dynamic equilibrium between immune and endocrine function can be altered not only by illness but by changes in lifestyle including caloric intake. The Comprehensive Assessment of the Long-term Effects of Reducing Intake of Energy Study (CALERIE™) is a large multi-year longitudinal study funded by the National Institute on Aging investigating what metabolic adaptations would result from a 25% caloric restriction sustained for 2 years and how these might positively affect primary aging and inform on aging biomarkers and chronic diseases of aging. A broad range of behavioral and physiological markers were measured at different levels of biology, making comprehensive integration and interpretation of data challenging. Here we propose a knowledge-based framework for integrating data that fully applies our current understanding of biobehavioral regulatory biology. Leveraging the Elsevier Knowledge Graph and the MedScan Natural Language Processing (NLP) engine we extracted from peer-reviewed literature 816 documented regulatory relationships linking a subset of 85 data elements readily available, including ~ 15 immune inflammatory and 12 endocrine markers. With a connection density of ~11%, this network was supported by > 23,000 citations or a median of 7 citations per regulatory interaction. A structural analysis of connectivity patterns identified inflammatory marker CCL2, insulin and testosterone as key mediators of large-scale coordinated response to caloric restriction. Ongoing analysis will identify immune and endocrine decisional logic that drives network dynamics, providing candidate intervention targets potentiating caloric restriction that remain explainable in the context of established pathway biology.

Clinical and Economic Burden Among Patients with Chronic Lymphocytic Leukemia Using a Novel Real-World Patient-Centered Database

Introduction Chronic lymphocytic leukemia (CLL) is the most common leukemia in adults, with an estimated 18,740 new cases of CLL in 2023. This study describes current treatment patterns, healthcare resource use (HCRU) and patient voice in patients with CLL in the US using a novel patient-centered database. Methods A retrospective study using data from the Inspire Integrated Analytical Database (IIAD), which links Inspire patient voice with electronic health records (EHR) and medical and pharmacy claims through HIPAA-compliant tokenization. Patients worldwide with various conditions join Inspire's online health community and contribute their experience through online posts and survey participation (see Figure 1). Medical and pharmacy claims in IIAD range from 2013 to present, with EHR data ranging from 2010 to present. Inspire members were included in the study if they resided in the US and had at least two diagnoses for CLL (ICD-10 codes C91.10, C91.11 or C91.12) on at least two distinct dates that were between 30 and 60 days of each other. The first observed CLL diagnosis date was considered the index date. Patients with an index date on or before 6/30/2022 were included in the analysis. Baseline patient characteristics and HCRU were evaluated. CLL treatment was identified from medical and pharmacy claims. Online posts which CLL patients have viewed at least once were evaluated. Post topics are defined as named entities recognized by Inspire's Natural Language Processing (NLP) engine from all of the discussion posts written by any Inspire members. Descriptive statistics were reported. Results A total of 331 CLL patients were identified in the online community with linked claims and EHR data. Median follow-up time was 59 months from index date. The median age at diagnosis was 73 and 49% of patients were female. Patients were geographically represented across the US. 63% had ≥100 claims from a median of 25 care sites across 33 providers. During follow-up a median of 3 hospitalizations per patient were observed. 5% of patients experienced a Richter's transformation and 22% developed a secondary malignancy (lung, gastrointestinal or skin cancer) after their index CLL date. 104 patients (31%) of the cohort had available data for CLL treatment. Median time to first CLL treatment was 11.7 months (range: 1 day - 86.8 months). Ibrutinib was the most common medication (n = 43), followed by venetoclax (n = 30) and rituximab (n = 20). The 5 most common topics viewed by patients on the Inspire community website included: skin, side-effect, muscle, biopsy, and infection. Prior to starting therapy, searches on the Inspire platform included terms such as “imbruvica,” “calquence,” and “living with CLL.” Searches after CLL treatment revealed that patients broadened search terms to include other cancer types such as “pelvic cancer,” “liver metastasis” and “melanoma.” Conclusions This report utilized a unique dataset that integrates patient voice with EHR and claims data to describe clinical characteristics, and treatment patterns of CLL patients drawn from the patient-centered Inspire digital health community. The online population had a median age of 73 at diagnosis, consistent with the average age of CLL diagnosis (70 yrs.). Patients in our cohort initiated treatment approximately 1 year after CLL diagnosis. Online activities on the Inspire platform demonstrated that CLL patients are engaged in searching and reading discussion posts concerning the impact of side effects and treatments on quality of life, and emphasized emerging concerns about the potential impact of secondary malignancies. Future real-world studies need to integrate patient voice to fully understand the consequences of a cancer diagnosis.

A - 6 Applying Artificial Intelligence Language Models for Knowledge Integration of Neuroimaging, Health Behavioral Assessment, and Clinical Intervention: Potential Impact on Brain Health in Special Operators.

Specialized military groups like Explosive Ordnance Disposal (EOD) personnel can suffer allostatic overload leading to complex health outcomes characterized as Operator Syndrome (OS). Here we (1) evaluate Brodmann's Areas (BA) brain regions associated with OS-related behavioral health symptoms then (2) explore how DoD/VA-recommended clinical interventions relate to these brain regions using large-scale automated text mining. Neuroimaging reported by Brodmann's Areas and 9 behavioral symptom measures and constructs of interest were extracted from full-text peer-reviewed publications using both rule-based and generative AI natural language processing (NLP) engines. The relationships linking BA regions with behavioral symptoms and the 7 interventions were assessed, with a focus on post-traumatic stress disorder (PTSD) and traumatic brain injury (TBI) severity. The network connected behavioral health symptoms and 39 of 52 BA regions through 131 interactions. Co-occurring PTSD and TBI associations to BA (counting ≥3 citations) were detected with the precentral gyrus (BA4), premotor cortex and supplementary motor cortex (BA6), middle temporal gyrus (BA21), superior temporal gyrus (BA22), angular gyrus (BA39), and dorsolateral prefrontal cortex (BA46). Behavioral interventions of Cognitive Processing Therapy (CPT), Prolonged Exposure Therapy, Cognitive Behavioral Therapy were found to modulate BA 6, with CPT alone also modulating BA 46. Only the pharmacotherapy fluoxetine (SSRI) related to BA6. NLP offers a rapid means of exploring complex injuries in Veterans. Future investigations will test these models using acquired data from EOD veterans and specific neuroimaging modalities (e.g., BOLD fMRI) to determine how they might improve understanding of health sequelae of environmental exposures and inform intervention.

Natural language processing for automatic evaluation of free-text answers — a feasibility study based on the European Diploma in Radiology examination

BackgroundWritten medical examinations consist of multiple-choice questions and/or free-text answers. The latter require manual evaluation and rating, which is time-consuming and potentially error-prone. We tested whether natural language processing (NLP) can be used to automatically analyze free-text answers to support the review process.MethodsThe European Board of Radiology of the European Society of Radiology provided representative datasets comprising sample questions, answer keys, participant answers, and reviewer markings from European Diploma in Radiology examinations. Three free-text questions with the highest number of corresponding answers were selected: Questions 1 and 2 were “unstructured” and required a typical free-text answer whereas question 3 was “structured” and offered a selection of predefined wordings/phrases for participants to use in their free-text answer. The NLP engine was designed using word lists, rule-based synonyms, and decision tree learning based on the answer keys and its performance tested against the gold standard of reviewer markings.ResultsAfter implementing the NLP approach in Python, F1 scores were calculated as a measure of NLP performance: 0.26 (unstructured question 1, n = 96), 0.33 (unstructured question 2, n = 327), and 0.5 (more structured question, n = 111). The respective precision/recall values were 0.26/0.27, 0.4/0.32, and 0.62/0.55.ConclusionThis study showed the successful design of an NLP-based approach for automatic evaluation of free-text answers in the EDiR examination. Thus, as a future field of application, NLP could work as a decision-support system for reviewers and support the design of examinations being adjusted to the requirements of an automated, NLP-based review process.Clinical relevance statementNatural language processing can be successfully used to automatically evaluate free-text answers, performing better with more structured question-answer formats. Furthermore, this study provides a baseline for further work applying, e.g., more elaborated NLP approaches/large language models.Key points• Free-text answers require manual evaluation, which is time-consuming and potentially error-prone.• We developed a simple NLP-based approach — requiring only minimal effort/modeling — to automatically analyze and mark free-text answers.• Our NLP engine has the potential to support the manual evaluation process.• NLP performance is better on a more structured question-answer format.Graphical

Artificial intelligence and the end of bounded rationality: a new era in organizational decision making

PurposeThe purpose of this viewpoint article is to serve as a discussion starting point regarding organizational leadership’s increasing reliance on AI – in particular, how the technology is used as a supplemental tool for supporting rational decision-making. Practical implications and directions for further research are presented.Design/methodology/approachWith its inception in economics, the concept of rationality has a rich history across multiple research domains. Based on that literature, coupled with the recent advancements in AI, the paper asks: will AI afford organizational leadership the ability to move from making bounded rational decisions to making fully rational decisions? The paper only scratches the surface of such a large question; however, the goal is to start the discussion around the topic.FindingsWhile bounded rationality supports efficient decision-making, a complete understanding of any given decision is typically limited, and as a result, neither accuracy nor effectiveness is guaranteed. As AI systems grow in speed and accuracy, they should provide positive support for organizational leaders to make fully rational decisions. AI’s ability to collect and organize data, analyze it, and offer decision alternatives may help close the gap between bounded and rational decision-making.Originality/valueAlthough AI research is not new, the recent developments in natural language processing engines has rapidly brought about new possibilities for their use in rational decision-making in the business and organizational context. This is fertile ground for future research, particularly in the area of organizational decision-making.

Tracking persistent postoperative opioid use: a proof-of-concept study demonstrating a use case for natural language processing

BackgroundLarge language models have been gaining tremendous popularity since the introduction of ChatGPT in late 2022. Perioperative pain providers should leverage natural language processing (NLP) technology and explore pertinent use...

Artificial emotions for charity collection: A serial mediation through perceived anthropomorphism and social presence

Despite the broad application of chatbot agents in online interactions, an ongoing debate persists regarding their persuasive role and human-like emotional disclosure. Our study adds to this debate by exploring the effect of chatbot agents’ emotional disclosure on people’s willingness to donate to a charitable cause, and by examining individual and serial mediation between the main effects of perceived anthropomorphism and social presence. To this end, two types of artificial intelligence chatbot agents—one disclosing factual information and another disclosing human-like emotion—were developed and trained using Dialogflow, a natural language processing engine. A total of 619 US residents were recruited through Amazon Mechanical Turk, an online crowdsourcing platform. Of these, 593 participants completed the required conversation with either version of the chatbot agent (factual vs. emotional), as well as the survey questionnaire, and therefore, were included in the final analysis. The participants exhibited a higher willingness to donate when they interacted with a chatbot disclosing human-like emotions than when they were only exposed to factual information. Moreover, this study found both individual and serial mediating roles of perceived anthropomorphism and social presence. Concerning the implications, theoretically, this study adds to the understanding of applying the notion of human interaction to that involving humans and chatbots. Practically, our findings can be of great help in increasing willingness to donate thereby enhancing fund-raising activities.

Deployment of Real-time Natural Language Processing and Deep Learning Clinical Decision Support in the Electronic Health Record: Pipeline Implementation for an Opioid Misuse Screener in Hospitalized Adults.

The clinical narrative in electronic health records (EHRs) carries valuable information for predictive analytics; however, its free-text form is difficult to mine and analyze for clinical decision support (CDS). Large-scale clinical natural language processing (NLP) pipelines have focused on data warehouse applications for retrospective research efforts. There remains a paucity of evidence for implementing NLP pipelines at the bedside for health care delivery. We aimed to detail a hospital-wide, operational pipeline to implement a real-time NLP-driven CDS tool and describe a protocol for an implementation framework with a user-centered design of the CDS tool. The pipeline integrated a previously trained open-source convolutional neural network model for screening opioid misuse that leveraged EHR notes mapped to standardized medical vocabularies in the Unified Medical Language System. A sample of 100 adult encounters were reviewed by a physician informaticist for silent testing of the deep learning algorithm before deployment. An end user interview survey was developed to examine the user acceptability of a best practice alert (BPA) to provide the screening results with recommendations. The planned implementation also included a human-centered design with user feedback on the BPA, an implementation framework with cost-effectiveness, and a noninferiority patient outcome analysis plan. The pipeline was a reproducible workflow with a shared pseudocode for a cloud service to ingest, process, and store clinical notes as Health Level 7 messages from a major EHR vendor in an elastic cloud computing environment. Feature engineering of the notes used an open-source NLP engine, and the features were fed into the deep learning algorithm, with the results returned as a BPA in the EHR. On-site silent testing of the deep learning algorithm demonstrated a sensitivity of 93% (95% CI 66%-99%) and specificity of 92% (95% CI 84%-96%), similar to published validation studies. Before deployment, approvals were received across hospital committees for inpatient operations. Five interviews were conducted; they informed the development of an educational flyer and further modified the BPA to exclude certain patients and allow the refusal of recommendations. The longest delay in pipeline development was because of cybersecurity approvals, especially because of the exchange of protected health information between the Microsoft (Microsoft Corp) and Epic (Epic Systems Corp) cloud vendors. In silent testing, the resultant pipeline provided a BPA to the bedside within minutes of a provider entering a note in the EHR. The components of the real-time NLP pipeline were detailed with open-source tools and pseudocode for other health systems to benchmark. The deployment of medical artificial intelligence systems in routine clinical care presents an important yet unfulfilled opportunity, and our protocol aimed to close the gap in the implementation of artificial intelligence-driven CDS. ClinicalTrials.gov NCT05745480; https://www.clinicaltrials.gov/ct2/show/NCT05745480.

Developing a Data and Analytics Platform to Enable a Breast Cancer Learning Health System at a Regional Cancer Center.

This study documents the creation of automated, longitudinal, and prospective data and analytics platform for breast cancer at a regional cancer center. This platform combines principles of data warehousing with natural language processing (NLP) to provide the integrated, timely, meaningful, high-quality, and actionable data required to establish a learning health system. Data from six hospital information systems and one external data source were integrated on a nightly basis by automated extract/transform/load jobs. Free-text clinical documentation was processed using a commercial NLP engine. The platform contains 141 data elements of 7,019 patients with newly diagnosed breast cancer who received care at our regional cancer center from January 1, 2014, to June 3, 2022. Daily updating of the database takes an average of 56 minutes. Evaluation of the tuning of NLP jobs found overall high performance, with an F1 of 1.0 for 19 variables, with a further 16 variables with an F1 of > 0.95. This study describes how data warehousing combined with NLP can be used to create a prospective data and analytics platform to enable a learning health system. Although upfront time investment required to create the platform was considerable, now that it has been developed, daily data processing is completed automatically in less than an hour.

Bio-SODA UX: enabling natural language question answering over knowledge graphs with user disambiguation

The problem of natural language processing over structured data has become a growing research field, both within the relational database and the Semantic Web community, with significant efforts involved in question answering over knowledge graphs (KGQA). However, many of these approaches are either specifically targeted at open-domain question answering using DBpedia, or require large training datasets to translate a natural language question to SPARQL in order to query the knowledge graph. Hence, these approaches often cannot be applied directly to complex scientific datasets where no prior training data is available. In this paper, we focus on the challenges of natural language processing over knowledge graphs of scientific datasets. In particular, we introduce Bio-SODA, a natural language processing engine that does not require training data in the form of question-answer pairs for generating SPARQL queries. Bio-SODA uses a generic graph-based approach for translating user questions to a ranked list of SPARQL candidate queries. Furthermore, Bio-SODA uses a novel ranking algorithm that includes node centrality as a measure of relevance for selecting the best SPARQL candidate query. Our experiments with real-world datasets across several scientific domains, including the official bioinformatics Question Answering over Linked Data (QALD) challenge, as well as the CORDIS dataset of European projects, show that Bio-SODA outperforms publicly available KGQA systems by an F1-score of least 20% and by an even higher factor on more complex bioinformatics datasets. Finally, we introduce Bio-SODA UX, a graphical user interface designed to assist users in the exploration of large knowledge graphs and in dynamically disambiguating natural language questions that target the data available in these graphs.

Identification of Preanesthetic History Elements by a Natural Language Processing Engine.

Methods that can automate, support, and streamline the preanesthesia evaluation process may improve resource utilization and efficiency. Natural language processing (NLP) involves the extraction of relevant information from unstructured text data. We describe the utilization of a clinical NLP pipeline intended to identify elements relevant to preoperative medical history by analyzing clinical notes. We hypothesize that the NLP pipeline would identify a significant portion of pertinent history captured by a perioperative provider. For each patient, we collected all pertinent notes from the institution's electronic medical record that were available no later than 1 day before their preoperative anesthesia clinic appointment. Pertinent notes included free-text notes consisting of history and physical, consultation, outpatient, inpatient progress, and previous preanesthetic evaluation notes. The free-text notes were processed by a Named Entity Recognition pipeline, an NLP machine learning model trained to recognize and label spans of text that corresponded to medical concepts. These medical concepts were then mapped to a list of medical conditions that were of interest for a preanesthesia evaluation. For each condition, we calculated the percentage of time across all patients in which (1) the NLP pipeline and the anesthesiologist both captured the condition; (2) the NLP pipeline captured the condition but the anesthesiologist did not; and (3) the NLP pipeline did not capture the condition but the anesthesiologist did. A total of 93 patients were included in the NLP pipeline input. Free-text notes were extracted from the electronic medical record of these patients for a total of 9765 notes. The NLP pipeline and anesthesiologist agreed in 81.24% of instances on the presence or absence of a specific condition. The NLP pipeline identified information that was not noted by the anesthesiologist in 16.57% of instances and did not identify a condition that was noted by the anesthesiologist's review in 2.19% of instances. In this proof-of-concept study, we demonstrated that utilization of NLP produced an output that identified medical conditions relevant to preanesthetic evaluation from unstructured free-text input. Automation of risk stratification tools may provide clinical decision support or recommend additional preoperative testing or evaluation. Future studies are needed to integrate these tools into clinical workflows and validate its efficacy.

Evaluation of an automated artificial intelligence (AI)/ natural language processing (NLP) engine to match patients (pts) with advanced solid cancers to biomarker-driven early phase (EP) clinical trials.

e13513 Background: The increasing complexity of precision medicine in oncology creates significant challenges in identification of pts for biomarker-driven clinical trials. Worldwide expansion of genomic testing will identify more pts with potentially actionable alterations, but the manual task of matching pts to clinical trials requires laborious interpretation of fragmented medical information and knowledge of suitable trial eligibility criteria. Inspirata’s Trial Navigator (TN) solution uses AI and NLP to automatically extract relevant information from both patient medical records and trial protocols to identify appropriate trials. We conducted a retrospective study to evaluate the ability of TN to match pts referred to the Guy’s Hospital EP trials unit. Methods: Anonymised referral letters and genomic reports for pts referred from Jan 2019 to Dec 2020 were used. A subset was used to train TN to extract relevant variables for matching against eligibility criteria for all UK Experimental Cancer Medicine Centre (ECMC) Network trials open within 160 km of Guy’s Hospital at the time of referral, favouring biomarker-matched studies. Finally, TN was tasked with finding 5 ECMC trial matches for 40 referred pts. Matches were scored for eligibility and tumour and/ or biomarker suitability using a quality index (QI): 0, ineligible; 1, generic match; 2, tumour match; 3, biomarker match; 4, tumour and biomarker match. Results: Of 40 pts, most common tumour groups included gastrointestinal (55%), HNSCC (20%) and gynae (10%). TN was able to correctly identify at least 1 suitable trial (median = 4) for 39/40 pts (97%). Selection biomarkers were available for 27 pts, of whom 21 (78%) were matched to at least one appropriate biomarker-directed trial, 5 (12%) were matched to biomarker-agnostic trials and one was not correctly matched due to exclusion for progressive CNS metastases. Of 13 pts without biomarker information, 11 (84%) were matched to at least 1 tumour-appropriate trial and 2 to tumour-agnostic trials. Of all 200 trial matches, 134 (67%) were appropriately matched, of which 80 (59%) were external to Guy’s Cancer centre. Conclusions: TN successfully matched 90% of these preselected cancer pts to an appropriate biomarker- or tumour-directed trial, indicating potential to improve access and recruitment to early phase cancer trials, including those external to the primary treating site with which referrers may not be familiar. A prospective multicentre evaluation of TN is warranted to further define improvements in trial recruitment and efficiency.[Table: see text]

Language & Coding Creativity

Abstract Machines are gaining understanding of language at a very rapid pace. This achievement has given rise to a host of creative and business applications using natural language processing (NLP) engines, such as OpenAIs GPT-3. NLP applications do not simply change commerce and literature. They raise new questions about how human beings relate to machines and how that symbiosis of communication will evolve as the future rushes toward us.

Software architecture of the question-answering subsystem with elements of self-learning

Within the framework of this paper, the analysis of software systems of question-answering type and their basic architectures has been carried out. With the development of machine learning technologies, creation of natural language processing (NLP) engines, as well as the rising popularity of virtual personal assistant programs that use the capabilities of speech synthesis (text-to-speech), there is a growing need in developing question-answering systems which can provide personalized answers to users' questions. All modern cloud providers proposed frameworks for organization of question answering systems but still we have a problem with personalized dialogs. Personalization is very important, it can put forward additional demands to a question-answering system’s capabilities to take this information into account while processing users’ questions. Traditionally, a question-answering system (QAS) is developed in the form of an application that contains a knowledge base and a user interface, which provides a user with answers to questions, and a means of interaction with an expert. In this article we analyze modern approaches to architecture development and try to build system from the building blocks that already exist on the market. Main criteria for the NLP modules were: support of the Ukrainian language, natural language understanding, functions of automatic definition of entities (attributes), ability to construct a dialogue flow, quality and completeness of documentation, API capabilities and integration with external systems, possibilities of external knowledge bases integration After provided analyses article propose the detailed architecture of the question-answering subsystem with elements of self-learning in the Ukrainian language. In the work you can find detailed description of main semantic components of the system (architecture components)

COVID-19 ChatBot

Abstract: Chat bots are software applications that help users to communicate with the machine and get the required result, this is where Natural Language Processing (NLP) comes into the picture. Natural language processing is based on deep learning that enables computers to acquire meaning from inputs given by the users. Natural language processing techniques can make possible the use of natural language to express ideas, thus drastically increasing accessibility. NLP engines rely on the elements of intent, utterance, entity, context, and session. Here in this project, we will be using Deep learning techniques which will be trained on the dataset which contains categories, patterns, and responses. Long Short-Term Memory (LSTM) is a Recurrent Neural Network that is capable of learning order dependence in sequence prediction problems. One of the most popular RNN approaches is LSTM to identify and control a dynamic system. We use an RNN to classify the category user’s message belongs to and then will give a response from the list of responses. Keywords: NLP – Natural Language Processing, LSTM – Long Short Term Memory, RNN – Recurrent Neural Networks.

Natural language processing for the surveillance of postoperative venous thromboembolism

BackgroundThe objective of this study was to develop a portal natural language processing approach to aid in the identification of postoperative venous thromboembolism events from free-text clinical notes. MethodsWe abstracted clinical notes from 25,494 operative events from 2 independent health care systems. A venous thromboembolism detected as part of the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) was used as the reference standard. A natural language processing engine, easy clinical information extractor-pulmonary embolism/deep vein thrombosis (EasyCIE-PEDVT), was trained to detect pulmonary embolism and deep vein thrombosis from clinical notes. International Classification of Diseases (ICD) discharge diagnosis codes for venous thromboembolism were used as baseline comparators. The classification performance of EasyCIE-PEDVT was compared with International Classification of Diseases codes using sensitivity, specificity, area under the receiver operating characteristic curve, using an internal and external validation cohort. ResultsTo detect pulmonary embolism, EasyCIE-PEDVT had a sensitivity of 0.714 and 0.815 in internal and external validation, respectively. To detect deep vein thrombosis, EasyCIE-PEDVT had a sensitivity of 0.846 and 0.849 in internal and external validation, respectively. EasyCIE-PEDVT had significantly higher discrimination for deep vein thrombosis compared with International Classification of Diseases codes in internal validation (area under the receiver operating characteristic curve: 0.920 vs 0.761; P < .001) and external validation (area under the receiver operating characteristic curve: 0.921 vs 0.794; P < .001). There was no significant difference in the discrimination for pulmonary embolism between EasyCIE-PEDVT and ICD codes. ConclusionAccurate surveillance of postoperative venous thromboembolism may be achieved using natural language processing on clinical notes in 2 independent health care systems. These findings suggest natural language processing may augment manual chart abstraction for large registries such as NSQIP.

Exploring Public Opinions on Renewable Energy by Using Conventional Methods and Social Media Analysis

Renewable energy is one of the most discussed topics nowadays, more so than traditional combustible energy sources and nuclear energy. There is an apparent need to study its public acceptance and the general public’s knowledge about it, especially in the context of more advanced inventions, such as smart grids, energy storage, or photovoltaic inverters. In this research, the results of a custom survey and those of several types of public social media datasets were processed to gain insights into the acceptance and awareness of these technologies. The second aim was to investigate technologies and related public opinions about manufacturers. One social media dataset was processed by a novel technology based on automatic data collection from popular social media sources, blogs, forums and public news. All the collected and anonymized data were restored in a custom-made data lake (unstructured database) and analyzed by a natural language processing (NLP) engine to extract the required information. The results showed that the Hungarian public’s knowledge differs from the European reference values by a very high percentage. A further goal for new research in the investigated area will be to achieve a business intelligence rated system in order to predict future trends and events. Since different methods provide results from different viewpoints, the conclusion is that combined analyses have to be used for investigating popular topics related to energy, such as various technologies, issues related to energy transition, or even car sharing services for a sustainable future.